Goldman says these banks and other ordinary companies may benefit most from next phase of AI trade

As mega-cap AI names wobble amid valuation and inflation worries, Goldman Sachs is highlighting a differentiated trade: companies poised to reap productivity gains from corporate AI adoption rather than the AI-infrastructure providers. Using a screen of the top 25% of Russell 1000 firms by share of wages susceptible to automation and high labor-to-sales ratios, Goldman assembled a basket that includes Bank of America, KeyCorp, PNC, Affirm, IBM and Zillow; the basket has returned 16% since Dec. 2023 (versus 23% for equal-weight S&P 500) but has outpaced earnings, prompting Goldman to call it an attractive risk-reward to broaden AI exposure. Goldman points to rising Q3 AI capex estimates and concrete examples—financial firms with >1/3 of wages exposed, BofA’s planned $4bn tech spend and potential 22% EPS lift, IBM’s HR automation and Zillow’s new AI tools—as evidence these companies can deliver efficiency-driven earnings upside beyond the AI infrastructure complex.

Analysis

Equity markets have softened amid valuation concerns, persistent inflation and worries about an AI bubble: the Dow and S&P 500 fell a fourth consecutive day and leading Nasdaq AI names such as Meta, Microsoft and Nvidia weakened this week, producing a cautious market tone despite a mildly positive sentiment score. Goldman Sachs highlights that these macro headwinds have increased investor focus on which companies beyond AI infrastructure can capture efficiency gains.

Goldman’s actionable screen targets the top 25% of Russell 1000 firms by the share of wages susceptible to AI automation and high labor-to-sales ratios, producing a basket that includes Bank of America, KeyCorp, PNC, Affirm, IBM and Zillow. That basket has returned 16% since December 2023 versus 23% for the equal-weight S&P 500, yet Goldman notes the stocks have lagged their earnings trajectory and Q3 pushed AI capex estimates higher, suggesting potential earnings tailwinds.

Concrete examples underpin the thesis: several financials have over one-third of wage costs exposed to automation, Goldman estimates Bank of America could see a 22% EPS lift and BofA plans $4 billion in tech spend, IBM has eliminated 200 HR roles via AI and Zillow launched AI-powered Zillow Pro. The opportunity is therefore productivity-driven earnings upside, but investors should weigh it against ongoing AI-infrastructure stress and broader valuation volatility.

Key Decisions for Investors

• Consider selective exposure to Goldman’s "AI Productivity Beneficiaries" (financials such as BAC, KEY, PNC, AFRM and tech-enabled names like IBM and Zillow) to capture potential productivity-driven EPS upside, sizing positions modestly given overall market volatility
• Use upcoming earnings and AI-capex disclosures as activation triggers—increase exposure only after confirmation of realized cost savings or upward EPS revisions (monitor BofA’s $4bn tech spend execution and IBM/Zillow operational milestones)
• Tilt away from concentrated AI-infrastructure longs until valuation and sentiment stabilize, or implement downside protection (options or hedges) because near-term macro/valuation risks could continue to drive sector volatility